Prof. James Gibson, University of Richmond School of Law
November 20, 2013
It’s back: Google Books, our favorite topic in this Intellectual Property Viewpoints series. Google Books is the project through which Google has brought its search capability to the text of more than 20 millions books (with the number still growing). To do so, Google must scan the book and convert the scan to readable text. And there’s the problem: Scanning is copying, and copying raises the question of whether the massive project is a massive violation of the copyrights in all those books. Thus the Authors Guild v. Google class action brought by authors and publishers against Google in a federal district court in New York.
Our earlier discussions of the case include antitrust and procedural problems with a proposed – but ultimately rejected – settlement, concerns about the direction the controversy might take in the aftermath of that rejection, musings about how the whole thing might be resolved by Congress or a public commission, and evaluations of Google’s chances on prevailing on a fair use argument, both in the abstract and in light of a ruling in the related Authors Guild v. Hathitrust case involving university libraries that provided books to Google for scanning.
Now the case has finally produced a ruling on fair use, and it is a clean win for Google. The decision has already generated a huge volume of commentary, and will probably continue to do so both here and elsewhere. In this essay, however, I thought I would start at the most basic level of analysis: the black-letter law that underlies the judge’s decision.
Although fair use has a long and storied history, its modern statutory incarnation lists four factors that inform the analysis: (1) the purpose and character of the defendant’s use, (2) the nature of the plaintiff’s copyrighted work, (3) the amount and substantiality of the portion the defendant used, and (4) the effect of the defendant’s use upon the potential market for the plaintiff’s work. Although all four factors should be considered, one lesson from earlier copyright cases is that it’s not just a matter of math; one party can prevail on three factors and yet lose the case. Moreover, the factors are interdependent, such that a court’s analysis of one factor can significantly influence the analysis of the others.
A prime example of this interdependence is Campbell v. Acuff-Rose, the Supreme Court’s 1994 decision that held, or at least strongly suggested, that rap group 2 Live Crew’s parody of Roy Orbison’s “Oh, Pretty Woman” was a fair use. In that case, the Court held that the purpose of 2 Live Crew’s use (the first factor) was parody, and that parody was a favored use in fair use analysis for various reasons, including that a parody transforms the underlying original rather than supersedes it. This holding colored the inquiry into the rest of the factors; in essence, it excused the fact that the original song was highly original (the second factor), that quite a lot of that originality was copied (the third factor), and that there might be a market for parodies (the fourth factor – ask Weird Al Yankovic).
A similar dynamic informed the Google Books decision. Relying on Campbell, the court first focused on whether Google’s use was transformative, and concluded that it was. “Google Books digitizes books and transforms expressive text into a comprehensive word index that helps readers, scholars, researchers, and others find books.” Of particular interest is the court’s conclusion that the use was transformative even though the expression in the copyrighted works was not itself transformed. This makes the Google Books decision different from Campbell, but here the court could cite a 2007 Second Circuit case, Perfect 10 v. Amazon.com, that held Google’s image search to be a fair use because it used the copyrighted images to facilitate research (a historically favored fair use) and information location rather than to supersede the expressive purpose of the originals.
The court’s decision that Google uses books in a highly transformative way essentially decided the case. It led the court to discount the fact that Google is a commercial enterprise – another first-factor consideration, and one that distinguishes Google from the defendants in the Hathitrust litigation. And the court then proceeded to dispose of the remaining three factors in fewer pages than it spent on the first alone. It blew through the second factor in five sentences, finding that it favored Google because most of the books are non-fiction. (One might think that this finding might dictate a different fair use outcome for some books over others, but instead the court found that it was a reason to bless Google Books as a whole. Again we see the overall transformative purpose – the utility of a comprehensive searchable database – affect the other factors.) The third factor received equally cursory treatment; indeed, the court did not even cite the cases (e.g., Sega v. Accolade) that best support the proposition that wholesale copying can be fair use as long as the defendant’s ultimate use reveals little or none of the plaintiff’s copyrighted expression. And discussion of the final factor, which is usually the most weighty of the four, was limited to the question of whether Google Books serves as a market substitute for reading the plaintiffs’ books. (It does not, but that simplistic fact alone is not enough to tell us whether other markets for the books – such as licensing markets – are similarly unaffected.)
That said, the ultimate outcome here is probably the correct one, and it’s certainly better than the outcome we would have had if the court had approved the proposed settlement back in 2011. For one thing, a fair use ruling allows others to compete with Google, whereas the settlement would have granted Google, and Google alone, the necessary licenses. Of course, even with fair use’s aegis, the up-front costs of creating a competing database are daunting, such that Google may have a natural monopoly. But who is to say that the libraries that gave Google the books in the first place – and that now have digital copies of the scanned texts themselves – won’t be willing to share them with a second comer?
So Google has gone a long way to winning the case on fair use grounds. Game and set to Google. But not quite match. The plaintiffs have already announced their plans to appeal to the Second Circuit. Their odds may not be good; the appeals court has previously hinted that it’s open to the fair use argument. (Indeed, the judge who issued last week’s fair use ruling is actually an appeals judge sitting by designation in the district court – Denny Chin, an increasingly influential jurist on cutting-edge copyright cases. He won’t hear the appeal of his own decision, of course.) Only one thing is certain: We are not done talking about this case yet.